Neural-network-based estimation of normal distributions in black-box optimization
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F22%3A10450929" target="_blank" >RIV/00216208:11320/22:10450929 - isvavai.cz</a>
Result on the web
<a href="https://doi.org/10.14428/esann/2022.ES2022-113" target="_blank" >https://doi.org/10.14428/esann/2022.ES2022-113</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.14428/esann/2022.ES2022-113" target="_blank" >10.14428/esann/2022.ES2022-113</a>
Alternative languages
Result language
angličtina
Original language name
Neural-network-based estimation of normal distributions in black-box optimization
Original language description
The paper presents a novel application of artificial neuralnetworks (ANNs) in the context of surrogate models for black-box opti-mization, i.e. optimization of objective functions that are accessed throughempirical evaluation. For active learning of surrogate models, a very im-portant role plays learning of multidimensional normal distributions, forwhich Gaussian processes (GPs) have been traditionally used. On theother hand, the research reported in this paper evaluated the applicabil-ity of two ANN-based methods to this end: combining GPs with ANNsand learning normal distributions with evidential ANNs. After methodssketch, the paper brings their comparison on a large collection of data fromsurrogate-assisted black-box optimization. It shows that combining GPsusing linear covariance functions with ANNs yields lower errors than theinvestigated methods of evidential learning.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
<a href="/en/project/LM2018131" target="_blank" >LM2018131: Czech National Infrastructure for Biological Data</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach
Others
Publication year
2022
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
ESANN 2022 proceedings
ISBN
978-2-87587-084-1
ISSN
—
e-ISSN
—
Number of pages
6
Pages from-to
187-192
Publisher name
i6doc.com
Place of publication
Belgium
Event location
Bruges, Belgium
Event date
Oct 5, 2022
Type of event by nationality
EUR - Evropská akce
UT code for WoS article
—